Stress analysis - contacts

[_TS__]

Hello,

I want to do a static analysis in order to see if an assembly can tolerate the designed force in a certain position. There is a structure which should have pinned contacts representing the ends of a cylinder and piston and one joint(see the picture). Should I manually define the joint surfaces' contacts as "sliding" or "separate" or something else? 

With joint contacts as "separate" or "sliding, no separation", the von Mises stress grows to a value which is ~30...33 times bigger than with bonded contacts. Obviously these(, some of the) contacts should not be bonded. Some messages are shown, too. Should I worry about a message "a model separates into 12 independent components"? (This message obviously doesn't occur if all the contacts are set to "bonded". Also note about soft spring doesn't come if all the contacts are bonded.) What to think about "soft spring was added"? Should I use the command "pin"?

Or maybe it would be best to set just the 2 joint surfaces which are not in the end of the cylinder+piston, just these 2 to something else than bonded?

I have read a bit from this conversation: Solved: Stress Analysis - Model separates into independent components - Autodesk Community - Invento...

Autodesk Inventor 2023
Thank you

[tobias]

Should I worry about a message "a model separates into 12 independent components"?

Yes, because the analysis will make no sense.

Big gaps you have to bond manually.

Your model is not bonded correctly. Every part you see moved from there place in the result you have to bond first.

[_TS__]

Thank you. Good to know this means the parts are not bonded. I was thinking it is allowed to get such messages when there are pinned connections.

Right now I have separated the part to 2 different calculations, "upper parts" from cylinder/piston = part 1 , and part2=parts which are below from cylinder/piston. Then I have added forces to ends where the cylinder/piston would be located and to the point where a pinned connection would be, and the forces are based on manual calculations. (Cylinder/piston is not analyzed anywhere since it is a simple thing to buy.) I am still thinking what would be the best way to place constraints to part1, if I am using this method where the original assembly-based part is divided into 2 analyses.